VK3/VC-037 and antenna tuning

I’ve known about this summit since Tony VK3CAT activated late last year and it looked like an easy one to get. Wanting to test out the amplifier and tune my ground plane it seemed a good choice for an afternoon activation. Unfortunately the forecast was for heavy rain starting from the afternoon, however by 2pm it was obvious this was mostly hanging to the west and only slowly approaching.

Following Glenn VK3YY’s as usual excellent directions I arrived just before the alerted time and setup. As it was a drive up, the laptop and scalar analyzer came with me and the following was produced initially:

Antenna was right at the bottom of the band edge which is what I remembered. After a bit of trimming I got the following (uncalibrated so off by 15dB or so):

1.5 swr across the band,  1.2 in the middle, much more suited for SSB operation in the upper half. Not too shabby. The radials will receive some attention next time.

Putting the 817 and amp on and things were very strange with a pronounced lack of band noise on 20m after I started. However I did manage to work S58AL,VK4FE, VK8GM, ZL1BYZ and ZL1TZW on 20 with good reports received but I struggled at times hearing the other side. When I switched off the amp after the last station the band noise immediately rose and a lightbulb lit up. The LED I added was drawing current through the relay coil so was preventing the RX/TX relay from releasing the contacts. Doh. Test activation well and truly justified, I never picked that up on the bench. 

Announcing I was listening on 144.150 I pulled down the ground plane and put the EFHW up for 40M. Contacts followed on 40m with VK2YK, VK5PAS, VK2FENG, VK5IS and VK2IO/M. No contacts made on 2m probably due to the last minute notice.

During this some rain arrived but not enough to pull the bothy bag out, I just covered the rig and relied on the waterproof paper of the logbook.

Leaving at 6:10 after a short second guess of my route, I was home by 7:30. Most of the drive was in heavy rain with plenty of water on the road in places. Well timed indeed.

Not a bad little summit, easy to get to, room for antennas, squidpole supports, a very scenic drive there and I qualified it with a 40 or 50dB attenuator on receive! Apologies to those who could hear this alligator but could not be heard themselves.


W6JL amp v1.1

My amp had an initial outing in a not totally finished configuration, several things were missing. 

In short:

  • Added a bicolor ON/TX led
  • Made the filters pluggable


20M LPF. Capacitors are 630v SMD.

The 20M filter shows enough attenuation by 20MHz to also be usable on 30M as expected and tests confirm this.

Done just in time for some small summit to summit party…

Only thing remaining is the SWR/display board.

40m LPF analysis

My W6JL amp didn’t seem to quite meet the spurious emissions limits in Australia with the 2nd harmonic a few dB short of 50dB down, though this was at the limits of my test equipment. This was a bit of a mystery as a 5 element LPF should have easiy taken care of the remaining 2nd harmonic.

Using the scalar analyzer neatly reveals why:

That little kink around 12Mhz is worth  5dB or so and so explains the gap.

The filters are being rebuilt as 7 element pluggable units and with hopefully better construction this will be the end of that.


Measuring the new 20m LPF i became suspicious of the results and it turns out the buffer amp in the analyzer is not ideal and is being affected by the load. 

Remeasured filter but with a 6dB attenuator after the amp:

Looks perfect.

I’ve now convinced myself that what i’m seeing of the 2nd harmonic is actually artifacts of the test setup. In any case, I’m moving to higher order LPF anyway.

Rebuilt the antenna analyzer as a scalar network analyzer

Some time back I built the k6bez antenna analyzer which utilized a cheap eBay DDS and and resistive bridge. It served it’s purpose though accuracy was questionable with no compensation of the diode voltage drop and lacking in power. My build being cheap and nasty lacked a case so it was fragile.

With an pressing need to build and tune some filters and wanting to put the thing in a box, it seemed like a good idea to rebuild it but as a Simple Scalar Network Analyzer allowing for more flexibility in measurements. I just happened to have a couple of AD8307 log detectors in the junk box having snapped them up at a good price a while back. The enclosures were also really well priced on eBay.

I’m fairly happy the way it turned out.   

I’ve set the amp so output power is 0dBm. The AD8307 log power detector is in a soon to be covered shielded enclosure, buffer amp below and the blue box contains a return loss bridge, thereby getting me back to where I started.  

The calibration results don’t seem too bad. The roll off at the lower end is likely due to either the 0.01uF cap on the detector input or the use of 43 material for the toroid in the bridge. It will do until the solar activity forces us onto 160m… Testing with a variety of different non-50ohm loads results in the expected return losses within fractions of a dB. The AD8307 is worth the cost.

Back to multibanding the skimmer and the hf amp…

Mt Sheridan VK4/NT-175

A family holiday to Port Douglas meant a chance to activate a new association. The superficial attractiveness of the NT region for SOTA is mitigated by the fact that most of the summits are in tropical rainforest. In the time Vk4 has been live, only 2 summits have been activated in NT.

A detailed searched revealed a few possible candidates, mostly around Cairns. One, Trintity Beach, had already been activated, but I found two that are mostly drive up and worth more. There were a couple of 10 pointers that might be  doable around Port Douglas but some research by Matt Vk1MA and his local connections helped eliminate these as easy time efficient options. Critical family holiday friendly attributes.

Anyway, VK4/NT-175 was picked as it had well detailed walking notes as it’s enroute to a well known, but apparently infrequently visited, lookout. In SOTA it’s known as Cairns 2, but the local name appears to be Mt Sheridan. 

Anyway, exiting highway 91 on Lake Morris Drive, take note of the odometer and just before 10km there is a hairpin with a grassy clearing where one can park. The below is looking back to where to you drive from.

Note that this road is a clearly marked cycling training ground and while I saw no insane people climbing at 3pm, it was quite busy at 5pm with both cyclists and walkers, so I suggest driving up in the heat of the day. The road is extremely narrow and windy, with several single lane sections. Expect to average 30km/h up. It would be fantastic on any form of bike, motorized or not.

On the way up I passed a sign marking the start of Dinden National Park confirming this would also count as vkff 0675. I was 99% sure before this was the case. Not that I actively participate in Parks, but useful for others. I will submit the log.

The access track to the comms facility on top is obvious just around the bend. Squidpole unfriendly gate included.

 It’s a 800m walk to the top but don’t fool yourself into thinking it’s that easy. It’s steep and you’ll likely  be climbing in 30degree heat with 90% humidity. Enough to get the heart racing and sweat pouring.

Part way up the nicely formed gravel track will fork. Correct choice is the  option on the left. The other leads to a dead end after quite a way.  

The summit:    

Running slightly late due to the need to stop and double check the map and turning right on the walk instead of left I was somewhat behind but quickly setup along the track. No problems and one would have enough space to put up an 80m antenna provided you can clear vegetation as no squidpole supports exist clear of it.

 Rig used was FT817 at 5w. Calling on 40m was fruitless, so switched to 20m where I had a nice stream of callers. I heard a reference to another parks station on 7.144, so qsyd there to try to work vk4ffab. Despite many tries and him being 59 to me, I was unheard. I did manage to pick up Fred Vk4FE in nearby Port Douglas though, and we had a decent chat. Fred had helped earlier with information regarding one of the other Summits and propagation. I may still get to Trinity Beach.

So 1 hour ten minutes of operation yielded 10 qso’s on 20m and 1 on 40m, all VK. I did change antennas and tried 15m for 10minutes but no luck. I was suffering high swr and no ATU. I really do need to tune that antenna. Like in VK6 and E5, a lot of CQing with silence and then a flurry of activity.

Given the propagation outlook with a geoeffecive coronal hole, things worked out quite well however QSB was wild and Europeans were reporting auroral effects. I was disappointed to have to shutdown as 20m was wide open to the US, but I had to be home in time for dinner. No time for dx cw. Dinner mission accomplished however.

CW Skimmer bliss

Doesn’t get much better than this:

That’s HDSDR running in parallel with SkimServ on 40m through the use of CWSL_Tee. The bottom 90khz of 40 is covered though I could extend this to 192k sampling. In theory I could also add the RttySkimmer to this, plus also other remote clients.

CW Skimmer experiments

Testing a Soft66RTL3 from Japan, basically a RTL SDR with a preamp, selectable band pass filters, and an up converter. It’s the little black box after the power injector for the active loop. 

The magic ingredient is the software on the raspberry pi, rtl_hpsdr written by N1GP, which makes up to 7 RTL devices look like a Hermes receiver. This then in turns allows skimmer server to be used which is optimised for multiband and low cpu usage. It runs with almost no cpu utilitistion in a virtualised Windows 10 environment, covering 192khz on 40m.

Seems to work, though I seem 15-20dB off VK4CT with the loop. Still working on calibrating it and ensuring it’s stable, though so far it looks good.

I also have a direct sampling RTL kit to finish to perform comparisons and the run two bands at once.

Another neat thing is with another dll, I might be able to run a websdr at the same time. Probably easier and better just to buy an elan receiver, but this is fun.